136 DIVISION II.—COURSE OF DEVELOPMENT OF FUNGI. 
The above phenomena close the cycle of development and in some species the 
course of development is actually limited to them; many observations at least have 
failed to detect anything further, for example in Pythium vexans and Artotrogus. 
We may therefore conclude that the essential points in the life-history of the whole 
group are confined to these phenomena. 
But in most species there is this difference, that the course of the development 
is extended by the intercalation of numerous propagative cells, gonidia. In some 
cases indeed the formation of gonidia is actually a necessary part of the entire 
development; the germ-tube which proceeds from the oospore developes into 
oF 

> FIG. 63. Cystopus candidus. A mycelium with young ia og. B jum og with oosp os and antheri- 
dium as. C mature ium og, os, D ripe pi in optical longitudinal section. Z, F, G formation of 
swarm-spores from oospores; Zendosporium. Magn. 400 times. 



a small rudimentary plant, which we may call a promycelium (see on page 
111), and this produces a few gonidia and then dies, while the gonidia give 
rise to new perfect fertile plants. This is the case with Phytophthora omnivora 
and Pythium proliferum which represent the intermediate mode of germination 
mentioned above. A very large majority of species, the two last-named among the 
number, form gonidia, not or not only in the way just described as terminal 
members of a short-lived alternate generation, but.as accessory products of every 
normally developed thallus; the gonidia are usually produced in such large quan- 
tities as to further the propagation of the species to an enormous extent by their 
germination, and they have such characteristic forms, that the characters of species, 
